Skip to content
Snippets Groups Projects
  1. Apr 26, 2019
    • eileen's avatar
      Add unit tests on contribution recur trxn_id, contribution recur processor id · 23faf53d
      eileen authored
      This is preliminary cleanup towards exposing cancel_reason as a search field, as committed to in proposal to
      add the field. This change
      
      - adds unit tests for the trxn_id & processor_id fields
      - makes the handling of them more generic
      - adds contribution recur field data to the contribution query object so it can be accessed.
      
      Some points about the last of these. Basically we have a situation where the query object needs metadata to
      avoid field by field handling. We have done this is some wierd & wonderful ways with dependencies on 'import' &
      'export' metadata info to the point where they were given to so many fields as to become meaningless.
      
      I think this method is better - ie. just adding the additional metadata to the query class for all the entities it
      deals with. A recently merged PR means 'where' data is still available using this method.
      
      Also I have added uniquenames to the contribution recur fields. Uniquenames no longer really matter outside
      - the query object
      - profiles
      - possibly import & export
      
      We only really expose recurring contribution info through the query object of these so it should be safe. In a future happy
      space apiv4 & namespacing will render uniquenames a distant memory.
      
      I couldn't rip out the special handling entirely without dealing with the labels that go into the qill. I figure
      we should merge this & then address that - but briefly the issue is that Matt has just changed all the labels
      to not quite work in this context so we need to discuss / agree how to manage 'Recurring Contribution Trxn ID' vs
      'Trxn ID'.
      
      I also didn't want to rip out the special handling fully in advance of https://github.com/civicrm/civicrm-core/pull/14118
      being merged as there could be a performance regression if we let more code hit that unnecessary join line
      23faf53d
  2. Apr 25, 2019
  3. Apr 24, 2019
  4. Apr 23, 2019
  5. Apr 22, 2019
  6. Apr 21, 2019
Loading